Methodological Validation and Clinical Usefulness of Carbon-14- Breath Test for Documentation of Presence and Eradication of

JoëlJ. Desroches, Raymond G. Lahaie, Michel Picard, Jacques Moráis,AndréDumont, Christiane Gaudreau, Daniel Picard and Raymonde Chartrand Departments of Nuclear Medicine, Gastroenterology, Pathology and Microbiology, Hôpital Saint-Luc, University of Montreal, Montreal, Canada

whereas the recurrence rate is only 10%-15% when a triple A simple [14C]urea breath test (C-14-UBT) was validated with aims antibiotic regimen against H. pylori is added (3,6). The Con of determining accuracy in documenting both the presence and sensus Development Conference of the National Institutes of proof of eradication of Helicobacter pylori infection. Methods: Fifty-six dyspeptic patients had with and C-14- Health (9) recommended that all patients with peptic ulcers who UBT. Eleven -proven H. py/or/-negative patients allowed are infected with H. pylori receive antimicrobial therapy. H. C-14-UBT normal value determination. Forty-three patients with pylori status determination has hence become a necessary tool recurrent and biopsy-proven H. pylori infection to guide the therapy of peptic ulcer disease. were included in an antimicrobial eradication protocol. Endoscopy Diagnosis of H. pylori infection can be reliably achieved with biopsies and C-14-UBT were done again 8 wk after initiation of during endoscopy by histologie analysis or culture of biopsy treatment in 35 patients. For C-14-UBT, 185 kBq (5 ßCt)of [14C]urea specimens (10-12). The unique -producing capability of was swallowed. Breath samples obtained up to 20 min were counted to calculate AS20, [(% 14CO2dose excreted/mmol of C02) these has also been used to detect them through urease testing of endoscopie biopsies. However, these procedures that x kg] at 20 min. Combined histologie and microbiologie analyses of require endoscopy were invasive, and simpler means of detec antral biopsies were used as a gold standard. Results: The positivity value was set as AS20 > 0.33% (mean + 3 s.d. of AS20 in H. tion were sought. While serology (ELISA) is simple, it cannot py/or/'-negative patients). Diagnosis of H. pylori infection was correct differentiate active from remote infection, nor can it quickly with C-14-UBT in 55/56 patients (44 true-positive, 11 true-negative document H. pylori eradication after treatment (3,6,10). Urea and 1 false-negative; sensitivity = 98%; specificity = 100%). As a breath testing was suggested as a promising noninvasive alter proof of eradication, C-14-UBT correctly classified 33/35 patients (5 native for these purposes (10,13-21). If H. pylori is present in true-positive, 28 true-negative and 2 false-positive; sensitivity = the , its urease will hydrolyze urea labeled with 100%; specificity = 93%). The C-14-UBT global performance carbon-13 or carbon-14 isotopes. The labeled H*CO3~ will be yielded sensitivity, specificity and accuracy of 98%, 95% and 97%, absorbed in the stomach and will diffuse in blood to be excreted respectively. A significant correlation (r = 0.84) was found between by the lungs as *CO2 that can be collected in breath samples. AS20 and the number of H. pylori colonies on culture. Conclusion: While both isotopes seem to offer similar diagnostic accuracy, This C-14-UBT is highly accurate both for diagnosis and proof of the 13C-urea breath test has the potential inconvenience of eradication of H. pylori infection and reflects the antral bacterial load. It is simple, fast and inexpensive, and it is therefore suitable for requiring more complex equipment, and it is more expensive. In clinical practice. addition, it usually requires the administration of a test meal and Key Words: Helicobacter pylori; urea breath test; carbon-urea; cold urea to the patient. This is not necessary with carbon-14, peptic ulcer disease; and the test is thus simpler, faster and more likely to achieve J NucíMed 1997; 38:1141-1145 clinical acceptance. This paper describes the use of a simple and quantitative MC-ureabreath test (C-14-UBT) in our institution. The aims of flelicobacter pylori was first identified and isolated by this study were: (a) to present a methodological validation of Marshall in 1983 (/). It is a gram-negative, urease-producing the test and determine normal values; (b) to determine its spiraled rod that has since been proven to be responsible for accuracy for detection of//, pylori infection; (c) to determine its type B chronic antral gastritis (2-8). H. pylori infection is now accuracy as a proof of//. pylori eradication after therapy; (d) to known to be the main cause of peptic ulcer disease and has been determine global performance of the test and (e) to show the associated with atrophie gastritis, gastric carcinoma and nonul- ability of the test to provide a semiquantitative evaluation of the cer dyspepsia (2-7). Recent studies have showed presence of//. antral bacterial load. pylori in more than 95% of duodenal ulcers and in 80%-95% of gastric ulcers not associated with nonsteroidal anti-inflamma MATERIALS AND METHODS tory drugs (2-4,6). The importance of//, pylori in the clinical Patients setting is related to ulcer recurrence. After I yr, between 55% For the methodological validation part of the study, 56 referred and 90% of ulcers recur when treated with anti-H2 therapy only, patients with persistent dyspepsia were recruited. After informed consent was obtained, all underwent upper gastrointestinal endos copy with antral biopsies as well as a '4C-urea breath test within 1 Received Apr. 12, 1996; revision accepted Sep. 19, 1996. For correspondence contact: JoëlJ. Desroches, MD, 950 Boulevard Mauricien, wk. H. pylori status was determined as described in the following Trois-Rivières-Ouest, Quebec, Canada, G9B 1V7. section, and the results of '4C-urea breath tests were then analyzed For reprints contact: Jacques Moráis,MD. Department of Nuclear Medicine, Hôpital Saint-Luc, 1058 Saint-Denis, Montreal, Quebec, Canada, H2X 3J4. against this gold standard in order to determine a cutoff value of

UREABREATHTESTINH. pylori INFECTION•Desroches et al. 1141 MOUTH PIECE UNIDIRECTIONAL VALVE

CONNECTING TUBE AIR OUTFLOW

AIR INFLOW

FIGURE 1. Apparatus for the 14C-urea CAP breath test. A disposable mouthpiece and unidirectional valve are linked to a connecting tube to a hole in the cap of a scintillation vial into which another hole is VIAL WITH pierced to allow air to be expelled. Vials TRAPPING SOLUTION with C0? trapping solution are succes sively attached to this cap. positivity. Evaluation of the accuracy of the l4C-urea breath test for Readings were done at 3 and 7 days. The bacterial load was detection of H. pylori infection was then performed. assessed semiquantitatively as 0 (no growth), 1+, 2+ or 3+ , Of the 56 patients, all those shown to be H. pv/or/'-positive by according to an estimate of the number of colonies on culture plate. analysis of antral biopsies performed during the upper gastrointes Our criteria for H. pylori positivity required the presence of at tinal endoscopy and known to have had at least one episode of least one of the following: (a) histologie identification of H. pvlori ulcers in the past were recruited for a H. pylori eradication or (b) microbiologie evidence of//, pylori on culture. A patient was protocol. Forty-three patients were thus randomized to one of three considered //. pv/or;'-negative when both criteria were absent. therapeutic regimens against H. pylori (22). Thirty-five patients These criteria we used are recognized in the literature as the gold had both an upper gastrointestinal endoscopy and a l4C-urea breath standard for H. pylori status determination. test done 8 wk after initiation of the 2-wk therapy against H. pylori. Carbon-14-Urea Breath Test Using the same cutoff value of positivity as above, these patients allowed evaluation of the performance of the '4C-urea breath test Materials. The apparatus is described in Figure I and consists of a 30-cm flexible plastic tube with, at one end, a mouthpiece as a proof of eradication of H. pylori infection. The global performance of our l4C-urea breath test (i.e., for connected to a unidirectional valve preventing any accidental aspiration of the trapping solution. The other end has an adapter diagnosis and eradication) was also assessed by analyzing the 91 linking the tube to a hole in the cap of a scintillation vial into which breath tests performed on the 56 patients. Finally, 48 of these another hole is pierced to allow air to be expelled when patients patients who had antral biopsies and cultures for H. pvlori were used to show the ability of our l4C-urea breath test to provide a breathe. The cap end allows vials to be attached for CO2 trapping. Each 25-ml vial contains 4 ml of benzethonium hydroxide (0.5 semiquantitative evaluation of the antral bacterial load. This was achieved by comparing the breath I4CO2 specific activity with a mmol/ml) in pure ethanol with lOOju.lof thymolphthalein (0.05%). semiquantitative evaluation (0-3 + ) of the number of H. pylori This bluish solution turns colorless when 2 mmol of CO2 are trapped in it. Four vials labeled TO, T5, TÕOand T20 were initially colonies on a culture plate. used for patients in our protocol (0-, 5-, 10- and 20-min breath Helicobacter pylori Status Determination samples). Carbon-14-urea is supplied as a freeze-dried solid, sealed Upper Gastrointestinal Endoscopy. During the standard upper under nitrogen in a borosilicate vial containing 9.25 MBq (250 gastrointestinal endoscopy, three or four antral biopsies were /iCi). This is dissolved in 50 ml of ethanol and stored in a obtained within 5 cm of the pylorus. Two were sent to pathology refrigerator. Each patient dose of 185 kBq (5 /j.Ci) is obtained by in formalin, one was sent to microbiology in an Eppendorf" pipetting 1 ml of the solution (185 kBq/ml). The three standard microfuge tube containing one drop of saline, and another (if vials, each containing 18.5 kBq (0.5 ju,Ci),are obtained by pipetting available) was used for rapid urease testing (CLOtest*). 0.1 ml. Each vial requires the addition of 15 ml of liquid scintillator Pathology. Tissue sections were stained with the Hematoxylin- after the CO2 collection, to allow beta scintillation counting in an Phloxin Safran technique. If needed, an additional argentaffin adequately calibrated instrument with quench correction. stain (Dieterle) was used in cases not obviously positive with the Patient Preparation. For every patient, weight and medications Hematoxylin-Phloxin-Safran technique. All specimens were re were recorded. Antacids and anti-H2s were stopped at least 12 hr viewed by the same experienced gastrointestinal pathologist and before the test while antibiotics and bismuth were stopped for 1 analyzed for presence of chronic active gastritis and histologie mo. History of gastric or abdominal surgery was noted. Patients identification of H. pylori. were asked to fast for at least 4 hr. Then, they were asked to rinse Microbiology'. Biopsy specimens were inoculated within the their mouth and gargle twice with water, which they then spit out. same working day over the surface of an agar supplemented with The test was explained and patients practiced by breathing until a Skirrow selective supplement and on an H. pylori medium (23). there was a color change in TO, which served as background. All agars were incubated at 35°Cunder microaerophilic conditions. Ingestion of '4C-urea then followed. The patients swallowed 25 ml

1142 THEJOURNALOFNUCLEARMEDICINE•Vol. 38 •No. 7 •July 1997 of water containing 185 kBq (5 /u.Ci)of '4C-urea, as fast as possible 10.00- with a straw, followed by another 25 ml of water. The test then proceeded with breath collections in vials labeled T5, TÕOand T20. Determination of I4CO2 Specific Activity at a Given Time ••• (AS,i,,J. Each of the four vials filled with 15 ml of liquid scintillator was counted once for 20 min. The equation providing 0) the percentage of the I4CO3 dose excreted per mmol of CO2 £;!•• collected in a vial at a given time and corrected for the patient's weight is: *••* AStimc = (% I4CO2doseexcreted/mmolofCO2) x kg 1.00 Ô net dpm sample T,,mc weight (kg) E dpm standard X 10 2 mmol of CÛ2 .33 where dpm sample Tljrnc= counts from sample at Ttlmc(time = 5. Ì 10 and 20 min); dpm sample TO = counts from sample at time 0 u (beginning: background); net dpm sample Tllmc = dpm sample Ttjmc —dpm sample TO; dpm standard = average of counts from l three standard vials: 10 = multiplication factor to correct for 1:10 i•e0.10- dilution of standard: and 2 mmol of CO2 = quantity of CO2 o" 000 collected in each vial. o The AS,,mc value has theoretical advantages over reporting a dpm value. First, it is less likely to be affected by small variations in the dose given to patients. Second, considering that the basal metabolic excretion of CO-, in man is 9 mmol/kg/hr, a heavier patient would exhale more CO2 for a same quantity of I4CO2, i.e., a lower specific activity. By providing correction for patient 0.01 weight, the AS,,,,,,,value can allow a semiquantitavive estimation of H. PYLORI H. PYLORI H. pylori urease activity. After beta scintillation counting, calcu NEGATIVE POSITIVE lations can be done readily, and results can be made available the (N=45) same day the test was performed. Dosimetri'. Dosimetry for the MC-urea breath test was described FIGURE 2. Scattergram of values of AS70 = [(% 14CO?dose excreted/mmol elsewhere (24,25) and is negligible, giving 0.10-0.20 /xGy/kBq of CO?) x kg] for H. py/ori-negati ve and H. py/on-positive patients, as (0.38-0.69 rad/mCi) to the bladder wall and an effective dose determined by combination of histologie and microbiologie criteria in 56 equivalent of 38-80 /iSv/MBq (0.14-0.30 mrem//xCi). For 185 patients. Cutoff value was set as 0.33%. kBq (5 /¿Ci),this represents 30-40 /j.Gy (3-4 mrad) to the bladder wall and an effective dose equivalent of 7-15 ju.Sv(0.7-1.5 mrem). allowed correct classification of patients as H. pylori-positive or H. pylori-negative in 55 of the 56 patients, with only one Statistical Analysis false-negative. To detect H. pylori infection, the sensitivity of The cutoff value for the MC-urea breath test was set as the mean the test was 44/45 = 98% (CI 95%, 94%-100%) and specificity + 3 s.d. of ASljrne in H. pylori-negative patients. The usual was 11/11 = 100% (95% CI, 94%-100%). The positive calculations of sensitivity, specificity, positive predictive value, predictive value was 100% (95% CI, 97%-100%) while the negative predictive value and accuracy were then performed. To negative predictive value was 92% (95% CI, 76%-100%); the assess correlation between variables for the semiquantitative eval accuracy was 98%. It is true that a receiver operating charac uation of the antral bacterial load, the Spearman regression analysis teristic analysis would have allowed a perfect discrimination of method was used. The 95% confidence intervals (95% CI) were H. pylori-positive and -negative patients with a 0.30% value of also calculated for the above values. AS20. Nevertheless, we preferred to choose a 0.33% cutoff value, knowing there had been preparation and technical prob RESULTS lems in the sole false-negative patient we obtained with this Methodological Validation of Carbon-14-Urea Breath Test cutoff value (see Discussion). The preliminary evaluation of data collected in the first part H. pylori Eradication Protocol of the study with 56 dyspeptic patients who had endoscopy and Of the 43 patients randomized in our eradication protocol, 35 a l4C-urea breath test revealed that samples obtained at 5 and IO patients were submitted to a post-treatment urea breath test and min were not giving any additional information when compared upper gastrointestinal endoscopy with biopsies. Of those 35 to 20-min samples. Data analysis was then performed only on patients, 28 patients with negative urea breath test all had 20-min samples with calculation of AS2(). Forty-five patients eradication of//, pylori (true-negatives). Of the seven patients were classified as H. pylori-positive and 11 patients were H. with a positive urea breath test, five were still infected with H. pylori-negative. For the H. pylori-negative group, statistics on AS2() showed a mean value of 0.125% with a 0.069% s.d. The TABLE 1 cutoff value for the 14C-urea breath test was determined to be Diagnostic Performance of Carbon-14-Urea Breath Test the mean + 3 s.d. of AS2(>in H. pylori-negative patients. A Breath test H. py/on-positive H. pylori-negative negative '4C-urea breath test thus corresponded to a value of Total AS2(, less than 0.33%. Figure 2 shows the AS2(, scattergram in PositiveNegativeTotal4414501111441256 the 56 patients and demonstrates the excellent discrimination achieved with a 0.33% cutoff. As shown in Table 1, this cutoff

URFABREATHTESTINH. pylori INFECTION•Desroches et al. 1143 TABLE 2 ASM»{(% "CO, dose excreted/mmol CO,) x kg) Performance of the Carbon-14-Urea Breath Test as Proof of H. pylon Eradication

Breath test H. py/ori-positive H. py/on-negative Total

Positive 2 7 Negative 28 28 Total 30 35 pylori (true-positives), and two patients did not present any residual infection on endoscopy (false-positives). Table 2 sum marizes the results obtained when the urea breath test was used as a proof of H. pylori eradication after antibiotic treatment. In this setting, our C-urea breath test presents a sensitivity of 100% (95% CI, 91%-100%), a specificity of 93% (95% CI, 84%-100%), a positive predictive value of 71% (95% CI, 37%-98%), a negative predictive value of 100% (95% CI, 1+ 2+ 3+ 96%-100%) and a 94% accuracy. H. PYLORI COLONIES ON CULTURE Global Performance of the 14C-urea breath test FIGURE 3. Values of AS20 = [(% 14CO2dose excreted/mmol of (Xy x kg], Table 3 provides the performance obtained by analysis of 91 expressed as mean ±s.d., as compared to a semiquantitative evaluation of breath tests in the 56 patients. Globally, our l4C-urea breath test the number of H. pylori colonies on culture. offered a sensitivity of 49/50 = 98% (95% CI, 94%-lOO%) and Our 14C-urea breath test protocol is interesting because of its a specificity of 39/41 = 95% (95% CI, 88%-100%). With a simplicity. First, it does not require any special oral preparation 55% prevalence of disease, the positive predictive value was 96% (95% CI, 90%-100%), the negative predictive value was like brushing teeth or antiseptic mouthwashing; it only requires gargling with water. Second, there is no need for test meal 98% (95% CI, 93%-100%) and accuracy was 97%. and/or cold urea administration, as usually required with 13C- Semiquantitative Evaluation of the Antrat Bacterial Load urea breath test. Third, only two breath samples need to be Figure 3 presents the comparison between the value of AS20 taken, one at TO (background) and the other at T20. This and the semiquantitative evaluation of the number of H. pylori compares favorably with multiple breath samples up to 60-120 colonies on culture on 48 antral biopsies performed with the min with some 13C-urea breath tests (18). Fourth, 14C-urea is same instrument. Regression analysis with the Spearman inexpensive and is detected by liquid scintillation counting, method demonstrated a very significant correlation (r = 0.84, which is already available in most hospitals (costs: SI3 Cana 95% CI, 0.72-0.91). Our MC-urea breath test thus provides an dian per patient for '4C-urea, other reagents and disposable estimation of the antral bacterial load. materials). Carbon-13-urea is expensive and traditionally re quires analysis by isotope ratio , an expen DISCUSSION sive apparatus with limited availability. These advantages of The relative merits of the various modalities to diagnose H. carbon-14 over carbon-13 largely compensate for the negligible pylori infection have been discussed (10,11,18). Many reports radiation burden of 185 kBq (5 /¿Ci)of 14C-urea, which is are now available to support the value of the urea breath tests approximately 50-100 times less than an upper gastrointestinal with either I3C or I4C as a noninvasive means of documenting series (16). both the presence and eradication of H. pylori infection (13- In our protocol, we chose to show results as AS20, although ¡8,20,21,26-30). It is also increasingly recognized now that the expressing results as raw dpm was shown to lead to approxi urea breath test will become the preferred method for the mately the same sensitivity and specificity for detection of //. assessment of the success of antibiotic therapy against H. pylori pylori infection (16). AS20 can be calculated in less than 1 min infection (6,18). Studies have already documented the excellent and presents theoretical advantages (see Materials and Meth prognostic value of a negative urea breath test post-treatment, ods). This value correlated well with a semiquantitative evalu with reinfection rates of 0.44%-2.2% per yr (6,31,32). Our ation of the antral bacterial load on culture in our study. Studies results further confirm the utility of a simplified 14C-urea breath have also demonstrated a similar correlation with the severity test in H. pylori infection. In this study, performance of our urea and activity of the gastritis (19). breath test was high both for diagnosis of H. pylori infection In our study, one infected patient had a false-negative urea and for documentation of H. pylori eradication. Globally, the breath test before antibiotic treatment with AS20 of 0.31%. For accuracy of our urea breath test was remarkable, with a 98% him. there was a small loss of liquid from the vial at the end of sensitivity and a 95% specificity. Our protocol thus compares the breath sampling, and he had been on omeprazole for more favorably with others (both with carbon-13 and carbon-14) than 3 wk until 12 hr before the test. At least one recent paper published so far showing a sensitivity of 90%-100% and a reported bacteriostatic and/or bactericidal effects of this medi specificity of 78%-100% (10,11,18). cation against H. pylori (33). These two factors probably contributed to decrease the AS20 slightly below the cutoff value. TABLE 3 Global Performance of the Carbon-14-Urea Breath Test Another patient with a history of partial and had a false-positive urea breath test with AS20 of Breath test H. py/ori-positive H. pylori-negative Total 0.43% after antibiotic treatment. It must be remembered that up to 3% of antral biopsy specimens are known to be false- PositiveNegativeTotal4915023941514091 negative when compared to urea breath test and serology (34). Moreover, patients with gastric surgery often have accelerated gastric emptying, and this could lead to the of urea

1144 THEJOURNALOFNUCLEARMEDICINE•Vol. 38 •No. 7 •July 1997 by urease-producing bacteria found in intestinal flora, thus 11. Lin SK, Lambert JR, Schembri M, et al. A comparison of diagnostic tests to determine falsely increasing AS2(). The last patient was also a false- Helicobacler pylori infection. J Gastroenlerol Hepalol 1992:7:203-209. 12. Lahaie RG, Boivin M, Dumont A, et al. Prevalence of H. pylori infection in patients positive after antibiotic treatment (AS20 = 1.44%) and was still undergoing upper G.I. endoscopy [Abstract]. Gastroenterology 1994:106:Al 16. 13. Marshall BJ, Surveyor I. Carbon-14 urea breath test for the diagnosis of Campy- significantly symptomatic. The discrepancy here could be lobacter pylori associated gastritis. J NucíMed 1988;29:11-16. related to the antro-fundal migration of H. pylori infection 14. Surveyor Õ,Goodwin CS, Mullan BP. et al. The l4C-urea breath test for the detection reported during treatment (33)\ antral biopsies would be nega of gastric Campylobacter pylori infection. Med J Aust l989;l51:435-439. 15. Rauws EAJ, Royen EAV, Langenberg W, Woensel JV, Vrij AA, Tytgat GN. '4C-urea tive, and the urea breath test would reflect the bacterial load breath test in C pylori gastritis. Gut 1989:30:798-803. now located in the fundus (34). Hence, in this case, it is very 16. Marshall BJ, Plankey MW. Hoffman SR, et al. A 20-minute breath test for likely that the urea breath test is a true-positive and that the Helicobacter pylori. Am J Gastroenterol 1991:86:438-445. 17. Henze E, Maffertheiner P. Clausen M, Burkhardt H, Adam WE. Validation of a antral biopsy is falsely negative. simplified carbon-14-urea breath test for routine use for detecting Helicobacter pylori Adequate patient preparation is important to get accurate noninvasively. J NucíMed 1990:31:1940 -1944. results with the l4C-urea breath test. Antacids should be stopped 18. Atherton JC, Spiller RC. The urea breath lest for Helicobacter pylori. Gut 1994:35: 723-725. for at least 12 hr (35). Antibiotics and bismuth should be 19. Debongnie JC, Pauwels S, Raat A, de Meeus Y, Haot J. Mainguet P. Quantification of stopped at least 4 wk before a diagnostic urea breath test and at Helicobacter pylori infection in gastritis and ulcer disease using a simple and rapid least 4-6 wk before documenting eradication of infection. carbon-14-urea breath test. J NucíMed 1991:32:1192-1198. 20. Graham DY, Klein PD. Evans DJ Jr. et al. Campylobacler pylori detected noninva Since omeprazole and sulfasalazine are now known to cause a sively by the 13C-urea breath test. Lancet 1987;i:l 174-1177. decrease in the bacterial load (33,36), it seems advisable to stop 21. Logan RPH, Dill S, Bauer FÉ,et al. The European 13C-urea breath test for the detection of Helicobacter pylori. Eur J Gastroenterol Hepatol 199I;3:915-92I. them before urea breath testing although the duration of 22. Lahaie RG, Lemoyne M, Poitras P, et al. A randomized trial of the efficacy of three abstinence is not well-documented yet. regimens for the eradication of Helicobacter pylori [Abstract]. Gastroenterology I995;108:A141. 23. Gaudreau C. Gilbert H. Lahaie RG. Comparison of Skirrow and Helicobacler pvlori CONCLUSION medium for the isolation of H. pylori [Abstract]. In: Newell DG. Ketley J. Fcldman This paper presented the methodological validation of a RA, eds. Eighth international workshop on campylobacters, helicobacters and related simple '4C-urea breath test protocol and its role in documenting organisms. Winchester: Central Veterinary Laboratory; 1995:6. 24. Stubbs JB, Marshall BJ. Radiation dose estimates for the carbon-14 urea breath test. the presence and eradication of H. pylori infection. Our 14C- J NucíMed l993;34:82l-825. urea breath test protocol, like others, is highly accurate both for 25. Munster DJ, Chapman BA, Burt MJ, et al. The fate of ingested 14C-urea in the breath test for Helicobacter pylori infection. Scand J Gastroenter 1993:28:661 666. diagnosis and for proof of eradication of//, pylori infection and 26. Raju GS. Smith MJ. Morton D, Bardhan KD. Mini-dose ( 1 (¿Ci)l4C-urea breath test provides a semiquantitative evaluation of the antral bacterial for the detection of Helicobacter pvlori. Am J Gastroenterol 1994:89:1027-1031. load. Our '4C-urea breath test is simple, fast and inexpensive, 27. Kao CH, Huang CK, Wang SJ, Hsu CY, Lin WY, Chen GH. Accuracy of a rapid 10-minute carbon-14 urea breath test for the diagnosis of Helicobacler pylori- and its radiation burden is negligible; it is therefore suitable for associated peptic ulcer disease. Eur J NucíMed I993;20:708-711. clinical practice. 28. Slomianski A, Schubert T, Cutler AF. 13C-urea breath test to confirm eradication of Helicobacler pylori. Am J Gastroenlerol 1995;90:224-226. 29. Braden B, Duan LP, Caspary WF, Lembcke B. More convenient 13C-urea breath test REFERENCES modifications still meet the criteria for valid diagnosis of Helicobacler pvlori infection. 1. Marshall BJ. Unidentified curved bacilla in gastric epithelium in active chronic Z Gastroenterol 1994:32:198-202. gastritis. Lancet 1983;i:1273-1275. 30. Klein PD, Graham DY. Minimum analysis requirements for the detection of Helico 2. Veldhuyzen van Zanten SJ, Sherman PM. Helicobacler pylori infection as a cause of bacter pylori infection by the 13C-urea breath test. Am J Gastroenterol I993;88:1865- gastritis, duodenal ulcer, gastric cancer and nonulcer dyspepsia: a systematic overview. 1869. Can MedAssocJ 1994;150:177-185. 31. Bell GD, Powell KU, Burridge SM, et al. Reinfection or recrudescence after apparently 3. Fennerty MB. Helicobacter pylori. Arch Intern Med 1994; 154:721-727. successful eradication of Helicobacter pylori infection: implications for treatment of 4. Peterson WL. Helicobacler pvlori and peptic ulcer disease. N Engl J Med 1991;324: patients with duodenal ulcer disease. Q J Med 1993:86:375-382. 1043-1048. 32. Borody TJ, Andrews P, Mancuso N, et al. Helicobacter pylori reinfection rate, in 5. Graham DY. Treatment of peptic ulcers caused by Helicobacter pvlori. N Engl J Med patients with cured duodenal ulcer. Am J Gastroenterol 1994;89:529-532. I993;328:349-350. 33. Logan RP, Walker MM, Misiewicz JJ, Gummett PA. Karim QN, Baron JH. Changes 6. Walsh JH. Peterson WL. The treatment of Helicobacter pylori infection in the in the intragastric distribution of Helicobacter pylori during treatment with omepra management of peptic ulcer disease. N Engl J Med 1995;333:984-991. zole. GUI 1995;36:12-16. 7. Matisko A, Thomson ABR. Helicobacter pvlori: beginning the second decade. Can J 34. Genta RM. Graham DY. Comparison of biopsy sites for the histopathologic diagnosis Gastroenterol 1995;9:333-343. of Helicobacler pylori: a topographic study of //. pylori density and distribution. 8. Thomson ABR. Helicobacler pylori and gastroduodenal pathology. Can J Gaslroen- Gastrointest Endose 1994;40:342-345. terol 1993;7:353-358. 35. Berstad K, Weberg R, Berstad A. Suppression of gastric urease activity by antacids. 9. National Institutes of Health Consensus Development Panel. Helicobacter pylori in Scand J Gastroenterol 1990;25:496-500. peptic ulcer disease. JAMA l994;272:65-69. 36. el-Omar E, Penman I, Cruikshank G, et al. Low prevalence of Helicobacter pylori in 10. Brown KE, Peura DA. Diagnosis of Helicobacter pylori infection. Gastroenterol Clin inflammatory bowel disease: association with sulphasalazine. Gut 1994:35:1385- North Am 1993;22:105-115. 1388.

UREABREATHTESTINH. pylori INFECTION•Desroches et al. 1145